AAS Worldwide Telescope: Seamless, Cross-Platform Data Visualization Engine for Astronomy Research, Education, and Democratizing Data

Home , Explorer 13, GitHub, Granat, Megan Schwamb

AAS WorldWide Telescope: Seamless, Cross-Platform Data Visualization Engine for Astronomy Research, Education, and Democratizing Data

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Rosenfield, Philip, Jonathan Fay, Ronald K Gilchrist, Chenzhou Cui, A. David Weigel, Thomas Robitaille, Oderah Justin Otor, and Alyssa Goodman. 2018. AAS WorldWide Telescope: Seamless, Cross-Platform Data Visualization Engine for Astronomy Research, Education, and Democratizing Data. The Astrophysical Journal: Supplement Series 236, no. 1.

Published Version https://iopscience-iop-org.ezp-prod1.hul.harvard.edu/ article/10.3847/1538-4365/aab776

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:41504669

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Open Access Policy Articles, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#OAP Draft version January 30, 2018 Typeset using LATEX twocolumn style in AASTeX62

AAS WorldWide Telescope: Seamless, Cross-Platform Data Visualization Engine for Astronomy Research, Education, and Democratizing Data

Philip Rosenfield,1 Jonathan Fay,1 Ronald K Gilchrist,1 Chenzhou Cui,2 A. David Weigel,3 Thomas Robitaille,4 Oderah Justin Otor,1 and Alyssa Goodman5

1American Astronomical Society 1667 K St NW Suite 800 Washington, DC 20006, USA 2National Astronomical Observatories, Chinese Academy of Sciences 20A Datun Road, Chaoyang District Beijing, 100012, China 3Christenberry Planetarium, Samford University 800 Lakeshore Drive Birmingham, AL 35229, USA 4Aperio Software Ltd. Headingley Enterprise and Arts Centre, Bennett Road Leeds, LS6 3HN, United Kingdom 5Harvard Smithsonian Center for Astrophysics 60 Garden St. Cambridge, MA 02138

ABSTRACT The American Astronomical Society’s WorldWide Telescope (WWT) project enables terabytes of astronomical images, data, and stories to be viewed and shared among researchers, exhibited in science museums, projected into full-dome immersive planetariums and virtual reality headsets, and taught in classrooms from middle school to college levels. We review the WWT ecosystem, how WWT has been used in the astronomical community, and comment on future directions.

Keywords: astronomical databases: atlases, astronomical databases: catalogs, astronomical databases: miscellaneous, astronomical databases: surveys, astronomical databases: virtual observatory tools, software

1. INTRODUCTION (AAS). Recent and near future AAS WWT development WorldWide Telescope1 (WWT) was first designed as a is focused on implementing operating system-agnostic tool for astrophysical data exploration and discovery in features to better enable astronomers to use WWT ei- the era of large and disparate data (Gray & Szalay 2004). ther individually or through archival data centers. WWT was an early attempt to answer an emerging ques- WWT is a scientific data visualization platform (see tion, “how will research techniques change in the era Figure1). It acts as a virtual sky, allowing users to explore all-sky surveys across the electromagnetic spec- arXiv:1801.09119v1 [astro-ph.IM] 27 Jan 2018 of multi-wavelength, multi-epoch, large datasets hosted throughout the Web?” Launched in 2008 by Microsoft trum, to overlay data from NASA’s Great Observato- Research, WWT was open-sourced (under the MIT li- ries, and import their own imagery and tabular data. cense) seven years later with copyrights transferred to In WWT’s 3D environment, users can explore plan- the .NET Foundation and project direction and manage- etary surfaces and elevation maps; orbital paths and ment taken up by the American Astronomical Society ephemerides of major and minor solar system bodies as well as asteroids; zoom out from the Solar System to view the positions and approximate colors of stars from Corresponding author: Philip Rosenfield the HIPPARCOS catalog (Perryman et al. 1997) and philip.rosenfi[email protected] galaxies of the Cosmic Evolution Survey dataset (COS- 1 http://worldwidetelescope.org MOS; Scoville et al. 2007). 2 Rosenfield et al.

WWT’s unique contextual narrative layer (Wong All of the code base is open-sourced under the MIT li- 2008) has set it apart from the myriad of other available cense and hosted on GitHub3. data visualization systems. Users can record or view recorded paths though the virtual environments and 2.1.1. Windows Application add narration, text, and imagery. With these “tours” The Microsoft Windows application contains the rich- users can share their stories, outreach specialists can est set of features and customizations. It currently create programming, and researchers can distill their serves as the basis of features that are ported to Web discoveries to other researchers or to the public. application. The WWT Windows application can run These features make WWT a seamless data visualiza- natively or through virtualization on either a local vir- tion program with an engaging learning environment. tual machine or on the Cloud. This combination has enabled WWT to deliver astro- Beyond the main tasks of visualizing the sky and 3D physical data hosted around the world not only to re- universe, the WWT Windows application features flexi- searchers, but to classrooms, planetariums, and educa- ble display modes, supports a variety of controllers, and tional outreach programs. Therefore, WWT is more can incorporate a wide variety of data. than a tool to aid astronomical research and discovery, it WWT can be displayed in power-walls (ultra high helps to democratize astronomical data and knowledge resolution displays created by connecting several moni- by allowing equal data access to all users. For exam- tors or projectors), 3D-stereo projections, virtual reality ple, only a few dozen content experts across academia headsets, interactive kiosk exhibits, and multi-channel and science centers have recorded tours curated within environments (e.g., in multiple projector full-dome plan- WWT or have organized WWT-related outreach efforts, etariums). Due to the high demand on computational yet WWT has reached millions2: from explorers using resources, there are limited plans to port these flexible the WWT application; to students visiting planetariums display modes to the Web application in the near future. and museums around the world; to students at all lev- Users may drive WWT with their keyboard and mouse els using WWT to augment their curriculum. WWT or customize a MIDI controller, an XBOX controller, or may be the only visualization software that when as- the Microsoft Kinect. tronomers add data and context, it becomes available A wide variety of data layers are both customizable to students–who may be geographically distant from a and controllable. Users can import 3D Models, tabular science center–to also investigate. data, orbital data, imagery (including all-sky maps and In what follows, we overview the technical aspects of planet surface maps), planetary digital elevation models, WWT in Section2, describe some major uses of this and catalogs. We discuss the WWT data files in more visualization tool in Section3, outline the open source detail in Section 2.5. management and sustainability plan in Section4, and conclude in Section5. 2.1.2. Web Application A growing subset of features from the WWT Windows 2. THE WORLDWIDE TELESCOPE ECOSYSTEM application are now delivered via a plug-in free browser- OVERVIEW based platform rendered by WebGL. The web application uses the same curated image and catalog data as In the nearly 10 years of WWT development, several the Windows application (see Section 2.5), it supports applications, services, and technologies were created in built-in tour authoring and playback, including a sim- order to deliver astronomical data from around the globe ple slide based or complex key-framed animations. The in a seamless visualization and narrative environment. web application can display tiled images, tours with au- In this section, we outline the main components of the dio and imagery, and most of the WWT intrinsic data WorldWide Telescope software ecosystem. layers. The WWT Web application leverages the Web Con- trol API for rendering. To better serve astronomical re- 2.1. Clients searchers, current development efforts lead by the Amer- WWT is offered as a stand alone Microsoft Windows ican Astronomical Society are focused on porting more application and Web-based applications and services. data layer features of the Windows application to the

2 Over 10 million unique Windows Application downloads from 3 See the WWT meta-repository: https://github.com/ 2007-2014 WorldWideTelescope/wwt-home WorldWide Telescope 3 web application and exposing them in the Web Control Show Image Service —enables an image to be shared by API (see Section 2.3). embedding information about the image in a URL, see the documentation for more details and Table2 for ex- 2.1.3. Extensibility of WWT Clients amples of its usage. Users can extend and customize their WWT client 2.3. APIs experience at a variety of experience levels. They can create a slide-based tour with linked slides (i.e, similar WWT has several application programming interfaces to a DVD menu). With the Windows application users (APIs) of interest to Web and Windows developers. can have full control over WWT tour rendering by using While all of the WWT code base is open sourced, it the time-line function (see this guide). They can add is incredibly complex, and above the software engineer- their own data layers (see Table1 for a list of supported ing level of most astronomers and astronomical software formats). And they can conﬁgure a planetarium with engineers. We expect most developers interested in ex- multiple computers and projectors to display WWT (see tending or customizing WWT to do so through the APIs this guide). Below we mention two more means to be (see Section4 for more discussion on code contribu- extend the WWT clients, by using “communities” and tions). by contributing documentation. 2.3.1. Windows APIs Communities —are user-uploaded WWT content (e.g., The Layer Control API (LCAPI) —allows local or remote tours and data layers) that are available to all users of applications to send data and control data layers in the the WWT clients though the Communities tab. Com- Windows application for integration in scenarios outside munities content can also be viewed on the WWT web- of the worldwidetelescope.org domain. There is deep in- site without launching the web application (see this tegration with the LCAPI and Microsoft Excel through guide). an Excel Add-In tied to the LCAPI, as well as python bindings (see Section 3.4). Documentation —All WWT documentation is now user contributed and maintained on GitHub4. Each guide Windows Client Socket API —is open-sourced, and used is rendered from Markdown as a GitBook5, which by WWT Kinect control to interact with WWT. is available to read online or download in PDF and 2.3.2. Web APIs eBook formats. There are also a growing number of user-contributed WWT tutorials on YouTube. WWT The Web Control API —is a complex and robust engine and the WWT Ambassadors (see Section 3.1) maintain that renders WWT objects and imagery into a canvas YouTube channels (WWT Videos, WWT Ambassadors element. Its purpose is to deeply integrate WWT vi- Videos). We further discuss contributing documentation sualization and tour playback into websites outside of in Section4. worldwidetelescope.org. It was written in C# and compiled into javascript using scriptsharp (See the API Doc- 2.2. Web Services umentation). Table3 lists examples of Web Control API usage. Embeddable Web Control6 —crosses the line between a ﬂexible web application and a software development kit Web Services API — Sets of hosted web services that de- (SDK). It implements a (growing) subset of the WWT liver or process data to and for the WWT clients and Web user interface to support visualizing data sets or other consumers, such as the ShowImage service refer- playing tours. It is designed to be cut-and-pasted into enced in Table2. Related documentation is hosted on websites outside of the worldwidetelescope.org domain. GitHub. The embeddable code-generation tool creates a WWT 2.4. Tools and SDKs namespace on the containing page and spawns an iframe which hosts the control. The webcontrol.aspx While part of Microsoft Research, WWT published script reads from the options in the containing div and software development kits (SDKs) to assist users and passes them to the iframe via a postMessage API (for developers with importing data (available here). These cross-domain scripting). tools are in the process of being converted to web services to support all operating systems.

4 See the WWT documentation meta-repository https://github. Sphere Toaster —takes small to medium sized images com/WorldWideTelescope/wwt-documentation. and outputs Tessellated Octahedral Adaptive Subdivi- 5 http://gitbook.com/ sion Transform (TOAST; McGlynn et al., in prep) tile 6 available at http://worldwidetelescope.org/Use/Embed pyramids, plate ﬁles, and WTML. 4 Rosenfield et al.

Table 1. Supported Data Formats

Name Description Reference or Documentation WTML (also XML-coded metadata description of how to find places and streaming Collections Guide called a “Collec- data services. tion File”) Tiled (quad tree) multi-resolution images sets in a variety of projec- Projection Refer- Tiled Data tions delivered through HTTP streams and described by metadata in a ence WTML file. Supported Projections 1. TOAST: Tessellated Octahedral Adaptive Subdivision Transform (TOAST; McGlynn et al., in prep) is a Hierarchical Triangular Mesh (HTM; Kunszt et al. 2000) based spherical projection without singularities at the poles and delivered through a quad-tree tile structure. 2. Mercator: Mercator-projected tiles (such as those provided by Bing and Google maps), for all-sky data, the poles are cut off. 3. Equirectangular (plate carrée):equidistant cylindrical projection for all-sky images, the poles have a singularity. 4. Tangential Study Maps: TAN-projected sections of small areas of the sky.

VAMP VAMP encoded images can be loaded and viewed directly, or tiled by Virtual Astron- our web service. omy Multimedia Project Stan- dards 3D models .3ds or .obj 3D file formats including textures and lighting. Guide for Adding 3D Models Tabular Data Either through cut/paste or import, tabular data can be visualized ... through an interactive user interface on the Windows application. WMS Web Mapping Service source for tiled and time-series data. Guide for adding WMS Data VO Table Tabular Data in VO Table Format can be viewed as tables or visualized VOTable Format Definition FITS Flexible Image Transport System (FITS) files in TAN-projection with FITS and WCS World Coordinate System (WCS) in celestial coordinates can be viewed Standard and stretched Shapefiles and Shape files can be loaded and viewed and converted to Well Known Text ESRI Shape- WKT (WKT) & Tables. WKT can be included in tabled for viewing complex file Technical geometry along with a text extension for displaying oriented text in 3D Description ODATA Tabular feeds from Open Data (ODATA) sources can be mapped in Open Data Speci- layers and dynamically refreshed at load time. fication TLE Two Line Element orbital data can be used for bulk display of orbits, or Guide for adding as a foundation for a reference frame. TLE Data Defini- tion of TLE SPICE (Acton NASA SPICE kernel data can be imported to display spacecraft or plan- Guide for adding 1996; Acton et al. etary emphemerides by minor manipulation within the NASA SPICE NASA SPICE 2017) GUI and Microsoft Excel into an .xyz text file. Data About SPICE WorldWide Telescope 5

Figure 1. Screen captures of the WorldWide Telescope User Interface. Clockwise from top left: “Sky” mode showing IRIS: Improved Reprocessing of Infrared Astronomy Satellite (IRAS) Survey (12, 60, and 100 µm (Red is IRIS100, Green IRIS60, Blue IRIS12) cross faded against 100 µm David J. Schlegel, Douglas P. Finkbeiner and Marc Davis (SFD) Dust Map (Data provided by IRAS and the Cosmic Background Explorer, Princeton University and University of California, Berkeley. Original IRAS data: NASA/JPL IPAC, IRIS Reprocessing: Canadian Institute for Theoretical Astrophysics/Institut d’Astrophysique Spatiale. TOAST-formatted mosaics were obtained using facilities of NASAs SkyView Virtual Telescope.); “Planet” mode showing Mars visible imagery (NASA/USGS/Malin Space Science Systems/JPL); “3D Solar System” mode (which includes HIPPARCHOS and COSMOS data) showing planets and their orbital paths, Oort cloud objects (white), Jupiter Trojans (yellow), and main belt asteroids (purple to white, colored by zone. Minor Planet ephemerides from the International Astronomical Union’s Minor Planet Center); and “Panorama” mode showing an image taken by Apollo 12 Lunar Module Pilot Alan Bean in 1969 and assembled at the NASA Ames Research Center in 2007. (Images courtesy of NASA and the Lunar and Planetary Institute).

Table 2. Show Image Service Usage Examples

Name, Website Instructions Montage Instructions on Using Montage to TOAST Images for WWT Chandra Photo Album Click on an image (from 2012 or earlier), click on the right panel “View on the Sky (WWT)” Hubble Space Telescope Click on an image and scroll to the bottom right “View in WorldWide Telescope.” Astrometry.net Click on or solve an image and scroll to the bottom right “View in WorldWide Telescope.” (Also available as a Flickr plug-in.) Spitzer IPAC AstroPix Click on an image and click “View in WorldWide Telescope” on the top right. 6 Rosenfield et al.

Table 3. Web Control API Examples

Name, Website Description Supplementary Code Import Image Import an AVM (Astronomy Visualization Metadata Standard) ImportImage.js tagged image or solve for AVM tags with astrometery.net and display it in the WWT Web Application Great Observatories NASA Great Observatories overview and image interactive Observatories.js Planet Explorer Travel from planet to planet in the 3D Solar System mode PlanetExplorer.js Spectroscopy Spectroscopy overview and interactive Spectroscopy.js Example Web Control API Usage Around the Web GLIMPSE 360 WWT “An infrared view of the disk, or plane, of the Milky Way galaxy ... Viewer assembled from more than 2 million snapshots taken over the past 10 years by NASA’s Spitzer Space Telescope.” Interactive Planck Viewer “...Compare the various components of the sky derived from ... Planck observations, as well as some other all-sky datasets.” ADS All Sky Survey “The ADS All-Sky Survey is an ongoing effort aimed at turning ADSASS WWT- the NASA Astrophysics Data System (ADS), widely known for frontend its unrivaled value as a literature resource for astronomers, into a data resource.” Milky Way 3D “A tool intended to organize and curate links to information ... about data sets relevant to our 3D understanding of the Milky Way” “’Chandra Source Catalog The Chandra Source Catalog (CSC) is ultimately intended to ... Release 2.0 be the definitive catalog of X-ray sources detected by the Chan- dra X-ray Observatory. To achieve that goal, the catalog will be released to the user community in a series of increments with increasing capability. The first official release of the CSC 2.0 will include 10,382 Chandra ACIS and HRC-I imaging observations released publicly through the end of 2014. When complete, CSC 2.0 will provide information for about 370,000 unique sources in several energy bands.” pyWWT The pyWWT package aims to make it easy to use WorldWide pyWWT on Telescope from Python (see Section 3.4). GitHub

Study Chopper —takes small to medium study images 2.5. Curated Data and outputs TAN-projected tile pyramids and WTML. The data shown WorldWide Telescope is a mix of curated data (stored on the Azure Cloud) and data dis- AVM Import Tool —takes Astronomy Visualization played from around the globe, including the National Metadata Standard-tagged images with WCS coordi- Virtual Observatory (NVO) registries. There are cur- nates and creates hosted tile pyramids and WTML ﬁles rently more than 90 all-sky surveys, 20 planet maps, to access them. This service is available online here. and 60 panoramas (see Appendix5: Tables4,5, and 6). In addition, WorldWide Telescope hosts over 500 Tile SDK —A software tool kit with samples that allows contributed images from education and public outreach users to create a custom data transformation pipeline oﬃces at Hubble, Chandra, Spitzer, Gemini, NOAO, to take images or elevation data from input format and NRAO, as well as from astrophotographers. output it to WWT compatible pyramids.

Terapixel Project —While not exactly an SDK, the Ter- 2.6. Supported Data Formats apixel Project is an open source pipeline that was used WWT supports a variety of data formats (see Figure to reprocess the original DSS data from scanned plates 2), which users can import via the Windows applica- into a Terapixel TOAST pyramid with globally opti- tion. Importing of data on the Web application is in mized color and brightness corrections (Agarwal et al. development, natively it currently supports FITS im- 2011). ages, and many more data formats are possible through WorldWide Telescope 7 an astropy-affiliated package (see Section 3.4). Table1 one or two full weeks of instruction from WWTA team lists current levels of data support. members (about Moon phases, seasons, and life in the Universe, see Udomprasert et al. 2014, 2016). They have 2.6.1. Plate Files pre-post assessment data showing outstanding learning The reasoning behind splitting large images into tiles gains from these interventions (Udomprasert et al. 2012, is to read from disk only as much as is absolutely nec- Udomprasert et al. in prep). Lesson plans and other essary for each database query (e.g., a view of a portion materials are available on their website. of the sky). The efficiency that tiling enables gives way WWT has also been used for Astronomy classes at in the limit of reading and transferring from the server, Beijing Shijia Primary School for three years. A Guide- many small files. The WWT team created “plate files” book on Interactive Astronomy Teaching (Primary as an efficient way to store, exchange and deliver tiled School Teacher version) designed by Chinese Virtual multi-resolution images. Plate files have an index and Observatory project is published by Popular Science data storage and allow millions of small files to be packed Press in Beijing. into a single file. Single large files copy up to 2 orders of 3.1.2. College Courses and Beyond magnitude faster than the millions of files they contain. There exist two plate file formats: Astronomy Courses —The WWTA program has helped to create WWT Windows Application-based introduc- 1. For densely packed complete surveys or studies tory astronomy labs on parallax and Hubble’s law. They with single generation data. have been used at Bucknell University and the Univer- 2. For sparse data that can contain multiple gener- sity of Massachusetts Amherst (see Ladd et al. 2015, and ations of data for a given tile index. This for- the WWTA Website). mat is great for building large surveys over time WWT has been part of introductory astronomy with streaming data, for example, the Mars HiRise courses at the Central China Normal University since data. 2009 (Qiao et al. 2010; Wang et al. 2015), and a graduate ISM course at Harvard University (Sanders et al. 2014). 3. WORLDWIDE TELESCOPE USAGE Co-author Weigel uses WWT to visualize spacecraft WWT has been used extensively at nearly all levels ephemerides and planetary texture maps in context in of astronomy education. For clarity, we separate the an immersive fulldome environment for the Introduc- following section by formal and informal education, and tion to Astronomy class at Samford University. These include education and public outreach (EPO) centers as visuals are presented and manipulated in real time in part of informal education. the Christenberry Planetarium on Samford’s campus.

3.1. Formal Education General Science Courses —Samford University makes further use of WWT in their Cultural Perspectives and Sci- The most prominent formal education program that entific Inquiry classes. Weigel teaches about the Scien- uses WWT in the WorldWide Telescope Ambassador tific Revolution using WWT in a planetarium. Students 7 Program (WWTA) which has evolved alongside WWT are guided on a journey from naked eye astronomical since 2010. observations and antiquated methods of scientific rea- The WorldWide Telescope Ambassadors Program soning and philosophy, through the age of the scientific aims to educate the public about Astronomy and Sci- revolution and the subsequent invention of telescopes, ence using WWT. It is run by a team of astronomers and finish with a look at current big data astronomy and educators at Harvard University, in collaboration and predictions for the future. with the AAS and Microsoft Research. They recruit and train volunteer ambassadors who help facilitate the Science Communication —The University of Washington use of WWT in educational settings like schools, science offered an undergraduate seminar on science communi- festivals, and museums. cation as applied to creating a tour in WWT (curriculum is available here). Students learned story boarding, dis- 3.1.1. Pre-College Courses tilling vs. dumbing down, avoiding jargon, and how to WWTA have directly served thousands of students in create WWT tours playable in the University of Wash- the classroom. Since 2015, more than 700 of these stu- ington planetarium. dents had extensive learning experiences that included 3.2. Informal Education WWT has shown to be an excellent data-driven as- 7 https://wwtambassadors.org tronomical education and public outreach environment. 8 Rosenfield et al.

Figure 2. Examples of supported data and models. Clockwise from top left: View of Earth with near Earth objects imported as Two Line Element (TLE) ﬁles (data from International Astronomical Union’s Minor Planet Center); 3D ISS Model (ISS Model from: Toshiyuki Takahe). A 3D rendering of a user-plotted catalog (The CfA Redshift Survey, green; Geller & Huchra 1989) over the COSMOS dataset (Scoville et al. 2007); Latitude, Longitude, and Altitude of a Mars Rover imported to WWT using the WWT-Excel plug-in (data courtesy of Joe Knapp, curiosityrover.com). For example, it inspired co-author Cui to initiate the IAU Working Group “Data driven Astronomical Edu- cation and Public Outreach” (DAEPO; Cui & Li 2018), it has been used to help capture and share indigenous Australian astronomical knowledge(Nakata et al. 2014), and has allowed planetariums to be ‘ﬂipped,’ allowing students to present planetarium shows to their peers.

3.2.1. Museums and Planetariums WWT has been used for live and recorded shows in major planetariums, such as the Adler Planetarium (Cosmic Wonder; Figure3) and the Morrison Planetar- ium (Earthquake). WWT has also proven to be an attractive choice for Figure 3. Adler Planetarium Audience viewing WWT dur- smaller planetariums, since the software is free and it ing Cosmic wonder credit: Microsoft Research Blog comes with a robust projector calibration system8. The following list of WWT-driven smaller planetariums is Mount Hood Community College, Central Washington incomplete, but includes the University of Washington, University, Samford University, and Bellevue College. In China, 6 WWT-driven planetariums have been built 8 See the Multi-Channel Dome Guide and 2 are currently under construction, with an Internet- WorldWide Telescope 9

Figure 5. WorldWide Telescope rendered in the full dome Christenberry Planetarium at Samford University. Credit: Figure 4. 2016 Science and Technology Week (China). Samford University Children are watching WWT tour shown inside a portable planetarium. based WWT-driven planetarium alliance for resource and experience exchange. Mobile, portable, or traveling planetarium programs have also begun to use WWT. For example, Discovery Dome9 packages WWT software with their inflatable domes, and the University of Washington, the Univer- sity of Oklahoma (as the Soonertarium), and Harvard University all have WWT-driven mobile planetariums programs. Nearly every mobile planetarium program has participated in some form of a local science festival, for example, Figure4 shows students watching a tour in a portable WWT planetarium. One of the unique uses of WWT in an educational set- ting comes from ‘flipping’ the classroom or in this case, the planetarium. Instead of students passively attending Figure 6. Samford University student experiencing VR over a lecture or recorded video in a planetarium, they can the surface of Pluto (New Horizons) in WWT. Credit: Sam- research and then create their own content to share in ford University a planetarium. This application of WWT has been in- dependently described by several groups. The DAEPO and present WWT tours for public audiences in a plan- has coordinated student authored WWT tour contests10 etarium (see Figure5). three times. The Cal Academy helped San Francisco The Christenberry Planetarium also uses WWT’s VR Bay Area students to present WWT in the Morrison capabilities in public outreach to promote the latest Planetarium (Roberts 2014). The University of Wash- astronomical discoveries and research by adding the ington works with Seattle-area middle and high schools “wow” factor by using cutting edge technology within to present WWT in a mobile planetarium (Rosenfield the university and the greater Birmingham, Alabama et al. 2014). The Christenberry Planetarium at Sam- community. Weigel and his students develop VR spe- ford University mentors middle and high school stu- cific WWT tours and host programs to both inspire and dents in summer programs, teaching them to produce teach youth in astronomy in Alabama (see Figure6). WWT has a kiosk mode where a user can interact with a guided tour or contained aspect of WWT. An in- 9 http://eplanetarium.com/software worldwide telescope.php complete list of museums and informational centers that 10 http://english.nao.cas.cn/ns/ConferenceNWorkshop/ have used WWT in this way include Harvard University 201706/t20170614 178121.html (see Figure7), the Adler Planetarium (see the Space Visualization Lab), and the Imiloa Astronomy Center 10 Rosenfield et al.

(e.g, LSST: Juri´cet al. 2015, and refs. therein). World- Wide Telescope was ﬁrst designed to exist in the post ftp-grep regime, and now that is has been open sourced and ported to the Web, it has plug and play potential for use in smaller-budgeted observatories and data archives.

Archival Centers —The next release of the WWT Web Control API will have data import/export, and selection features which will make it powerful link between data archives and researchers. Observatories and data centers can insert the web control on their website and have a robust, open source, data archive visualization engine. Observatories or data archive centers that use WWT in Figure 7. Harvard University’s WWT-driven Kiosk this way will automatically benefit their EPO efforts, since EPO staff would have the same access to data and be able to create tours or build a custom user interface of Hawai’i which had a Microsoft Kinect driven WWT over the WWT Web Control API. exhibit “The Universe at your Finger tips”. In addition, abstract services, such as ADS11 can use 3.2.2. Education and Outreach Centers the WWT interface as another means for users to dis- cover research in astronomy. The Zooniverse’s Astron- Both WWT clients have served as a conduit for pub- omy Rewind project12 is a citizen scientist project to lishing EPO imagery and stories. For example, the “Ex- place AAS journal figures dating back to the 1800s into plore” tab shows curated feeds from the Hubble Space the Astronomy Image Explorer13 with AVM tags so they Telescope, Spitzer Science Center, Chandra X-ray Ob- can be viewed in WWT. servatory, and the European Space Agency. Several outreach efforts have also made use of the WWT Web Con- Data Analysis Tools —Several data analysis tools are be- trol API (see bottom of Table3) and the WWT Show ing built off of the Web Control API. For example, a Image service (see Table2). link between JS914 and WWT has been established as 15 3.3. Astronomical Research part of the AAS Astrolabe project . Users of this service can upload their images to JS9, manipulate them WWT enters the astronomical research work-flow at as they would in DS9, and view the result in WWT. several points: discovery, helping visualize big and wide Python bindings for the WWT Web Control API are datasets; analysis, WWT APIs allow analysis tools to a new focus and were first prototyped in Glue-viz16. link data; and dissemination, WWT can enhance figures and presentations, and researchers can create video PyWWT: A Python interface to WorldWide Telescope — abstracts to accompany their submitted manuscripts to PyWWT is a package that provides access to much of journals (see Section 3.5). WWT and its APIs allow the Windows and web clients’ functionality to Python observatories and mission centers to adopt a common on a variety of platforms. It allows the WWT web visualization interface for their databases. In the fol- client to be used as an interactive widget inside Jupyter lowing section, we discuss the most recent efforts and Notebooks and and Jupyter Lab (pictured in Figure8; future outlook led by the AAS to help the astronomical Kluyver et al. 2016), as well as through a standalone in- research community benefit from WWT technologies.

3.4. Discovery & Analysis 11 http://adsabs.harvard.edu Astronomers interested in using survey data (e.g., 12 https://www.zooniverse.org/projects/zooniverse/ astronomy-rewind SDSS, LSST, Gaia, DESS) have most likely shifted from 13 http://www.astroexplorer.org the traditional ftp-grep data analysis model, where a 14 https://js9.si.edu; the online javascript version of DS9 (http: scientist ﬁrst downloads a copy of data (ftp) and then //ds9.si.edu) with a public API uses an analysis package to identify for patterns, or 15 The Astrolabe Project is creating a new open-access reposi- answer their research question (grep) (see e.g., Gray tory for previously un-curated astronomical datasets, building on existing CyVerse infrastructure with robust cloud-based resources & Szalay 2004). Instead, astronomers analyze their for managing, linking, processing and sharing research data. data by querying remote databases, and soon, access- 16 Glue-viz is Python library to explore relationships within and ing hosted computing facilities to perform their analysis among related http://glueviz.org/en/stable/, see WWT Plugin WorldWide Telescope 11

WorldWide Telescope is managed by the AAS its (MIT) copyright is held by the .NET Foundation. Below we overview the management and sustainability plan for this open source software project.

4.1. Project Leadership AAS Board of Trustees (AASBoT) is in charge of managing the AAS WWT project. The AASBoT has mandated the AAS WWT Steering Committee to con- tinually assess the WWT project against metrics based on the WWT’s purpose statement as well as the AAS’ purpose statement. The AAS WWT Steering Commit- tee and AAS WWT Project Director work together to Figure 8. A screen shot of pyWWT, a Python interface for interpret the purpose statements into initiatives with WorldWide Telescope based on the Web Control API. quantitative metrics for success. teractive Qt-based viewer/widget17, and also provides a 4.2. Open Source Contributions Python-based client for the WWT desktop application The longevity and success of open source software de- on Windows (using the LCAPI, see Section 2.3). pends on the community. We are striving to build a In all cases, users can pan and zoom with the mouse, inclusive WWT community. We expect everyone in the display multiple layers from all-sky surveys, and play WWT community to follow our code of conduct and a tours. PyWWT makes it easy to add annotations, or contributing guide. Below, we discuss two avenues of shapes, that can be used to highlight celestial objects, contribution: contributing to the code base, and con- enhance tours, or illustrate fields of view for telescopes, tributing documentation and examples. among many other applications. The package is written to integrate smoothly with Astropy(Astropy Collabora- 4.2.1. Code Contributions tion et al. 2013) for easy adaptability to typical research We maintain a meta-repository, wwt-home, which pri- work-flows such as those described in Section 3.4. Sup- marily contains a table of information on the different port for Astropy tables and displaying local FITS images WWT-related repositories in the WorldWideTelescope is forthcoming. GitHub organization. Importantly, each table entry in- PyWWT is under active development and is hosted cludes a description of the project as well as the skills under the WorldWideTelescope GitHub organization required to contribute. Until we have a robust user com- (see Table3), where its full code base is open-sourced munity to review code contributions (as well as propos- (version 0.3.0 was launched in December 2017). als for enhancements), they are considered by the WWT Advisory Board. 3.5. Dissemination We have seen examples of the WWT narrative layer 4.2.2. Documentation Contributions (tour features) in sharing stories and teaching astron- As discussed in Section 2.1.3, all of our documentation omy. WWT has also been used in video abstracts, which has been open sourced and hosted on GitHub, accessible have converted WWT tours into videos distilling a schol- via our meta-repository, wwt-documentation including arly publication (see the Publisher’s website of Berri- code samples. man & Good 2017; Batygin & Brown 2016; Yusef-Zadeh Contributers can edit existing documentation or cre- et al. 2015; Currie et al. 2015). Future WWT Web Con- ate their own, even if they are not familiar with GitHub trol development will allow for embedding WWT in 2- or Markdown. In addition, they can create documen- and 3D modes as an interactive figures in scholarly pub- tation within a GitBook or import files created using lications. Microsoft Word. Each new contribution will be added as a line to the wwt-documentation table which is the 4. MANAGEMENT AND SUSTAINABILITY OF basis of the www.worldwidetelescope.org’s User Guides WORLDWIDE TELESCOPE Page. Sample code can be published as gists and linked to our meta-repositories or as Jupyter notebooks (e.g., py- 17 https://www.qt.io WWT example notebooks). 12 Rosenfield et al.

We will continue to engage and consult with the analysis tools; c) interactive (or video) figures for broader astronomical community to plan new features scholarly publications and presentations. and services via AAS member polls, WWT online fo- rum, tracking GitHub issues, the WWT help desk Authors thank WorldWide Telescope Advisory Board ([email protected]), and feedback during workshops and whose effort of to support the WWT project since its webinars (direct and via assessments). inception and help transition WWT from Microsoft to 4.2.3. Contributing WWT Products the AAS insured its existence: Sarah Block, Andy Con- In the future, we hope to create a place on the Web nelly, Chen-Zhou Cui, Cristine Donnelly, Karl Fogel, to house WWT-related workshop materials, contributed Ron Gilchrist, Alyssa Goodman, Morgan Griffith, Bryan WWT lesson plans, 3D models, music, and more–that Heidorn, Robert Hurt, Jonathan Fay, Erin Johnson, Su- meet the existing skill sets of the contributers. sanna Kohler, Knut Olsen, Fred Rasio, Doug Roberts, Philip Rosenfield, Gretchen Stahlman, Mark SubbaRao, 5. CONCLUSIONS Julie Steffen, Matt Turk, Patricia Udomprasert, Jaap We have reviewed the software ecosystem of WWT Vreeling, A. David Weigel, Ryan Wyatt, Martin Wood- and many of its uses in astronomy. WWT is astronom- ward, and Curtis Wong. We also thank the the current ical visualization software what can directly link to all and past AAS WWT Steering Committee members for levels of education, in that way, WWT can democratize support, ideas, and feedback on development initiatives: astronomical data, but only with the help of the astro- Amanda Bauer, Nancy Brickhouse, Jack Burns, Buell nomical community. Jannuzi, Jessica Kirkpatrick, Sarah Loebman, Kevin In summary: Marvel, Megan Schwamb, and Chick Woodward. We thank Doug Roberts and Julie Steffen for AAS WWT • WWT is open source (MIT license) and managed project leadership. by the American Astronomical Society. Chen-Zhou Cui thanks support from National Nat- ural Science Foundation of China (NSFC)(11503051, • WWT has two main applications: 1) a Microsoft U1531111, U1531115, U1531246, U1731125, U1731243). Windows application that is a mainstay in plane- Chen-Zhou Cui also thanks the National R&D In- tariums and international education; and 2) a We- frastructure and Facility Development Program of bGL powered Web application that is operating China, “Earth System Science Data Sharing Platform” system agnostic which we expect most US-based and “Fundamental Science Data Sharing Platform” astronomers to prefer. (DKA2017-12-02-XX). • Each application has APIs and services to al- This publication makes use of data products from the low users and developers to customize and extend Two Micron All Sky Survey, which is a joint project WWT without needing the high level of C# pro- of the University of Massachusetts and the Infrared gramming knowledge to advance the core WWT Processing and Analysis Center/California Institute of components. Technology, funded by the National Aeronautics and Space Administration and the National Science Foun- • WWT has been used innovatively in formal and dation. informal educational settings, supplementing: a) We acknowledge the use of the Legacy Archive for early learners physical understanding of the sea- Microwave Background Data Analysis (LAMBDA), part sons and moon phases; b) college students’ intro- of the High Energy Astrophysics Science Archive Center ductory astronomy and general science courses as (HEASARC). HEASARC/LAMBDA is a service of the well as at least one graduate astronomy course; f) Astrophysics Science Division at the NASA Goddard planetarium education, e.g., by enabling students Space Flight Center. to present planetarium shows to their peers. Funding for the SDSS and SDSS-II has been provided • Educational programs have used WWT in a va- by the Alfred P. Sloan Foundation, the Participating riety of settings: computer labs, kiosks, virtual Institutions, the National Science Foundation, the U.S. reality headsets, and planetariums. Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, • Ongoing development of the WWT Web Control the Max Planck Society, and the Higher Education API will help to embed the powerful tool in astro- Funding Council for England. The SDSS Web Site is nomical research: a) as a visualization engine for http://www.sdss.org/. The SDSS is managed by the data archives; b) linked to python and javascript Astrophysical Research Consortium for the Participat- WorldWide Telescope 13 ing Institutions. The Participating Institutions are the versities for Research in Astronomy, Inc., under NASA American Museum of Natural History, Astrophysical In- contract NAS5-26555. Support for MAST for non-HST stitute Potsdam, University of Basel, University of Cam- data is provided by the NASA Office of Space Science bridge, Case Western Reserve University, University of via grant NNX13AC07G and by other grants and con- Chicago, Drexel University, Fermilab, the Institute for tracts. Advanced Study, the Japan Participation Group, Johns The National Radio Astronomy Observatory is a facil- Hopkins University, the Joint Institute for Nuclear As- ity of the National Science Foundation operated under trophysics, the Kavli Institute for Particle Astrophysics cooperative agreement by Associated Universities, Inc. and Cosmology, the Korean Scientist Group, the Chi- This research has made use of data obtained from or nese Academy of Sciences (LAMOST), Los Alamos Na- software provided by the US National Virtual Observa- tional Laboratory, the Max-Planck-Institute for As- tory, which is sponsored by the National Science Foun- tronomy (MPIA), the Max-Planck-Institute for Astro- dation. physics (MPA), New Mexico State University, Ohio We acknowledge the use of public data from the Swift State University, University of Pittsburgh, University data archive. of Portsmouth, Princeton University, the United States This research has made use of data and/or services Naval Observatory, and the University of Washington. provided by the International Astronomical Union’s Mi- The Digitized Sky Survey was produced at the Space nor Planet Center. Telescope Science Institute under U.S. Government This research uses services or data provided by the grant NAG W-2166. The images of these surveys are Science Data Archive at NOAO. NOAO is operated by based on photographic data obtained using the Oschin the Association of Universities for Research in Astron- Schmidt Telescope on Palomar Mountain and the UK omy (AURA), Inc. under a cooperative agreement with Schmidt Telescope. The plates were processed into the the National Science Foundation. present compressed digital form with the permission of This work made use of the IPython package (Pérez these institutions. & Granger 2007), Scikit-learn (McKinney 2010), and This research has made use of the USNO Image and Astropy, a community-developed core Python package Catalogue Archive operated by the United States Naval for Astronomy (Astropy Collaboration et al. 2013). Observatory, Flagstaff Station (http://www.nofs.navy. The acknowledgements were (partially) compiled us- mil/data/fchpix/) from Monet(1998). ing the Astronomy Acknowledgement Generator. Based on observations made with the NASA Galaxy Facility: CGRO, COBE, EUVE, Fermi, GALEX, Evolution Explorer. GALEX is operated for NASA by GRANAT, HEASARC, IRSA, MAST, NRAO, OCA:IRIS, the California Institute of Technology under NASA con- Planck, ROSAT, Sloan, Swift, VLA, WISE, WMAP, tract NAS5-98034. WSRT Some of the data presented in this paper were obtained from the Mikulski Archive for Space Telescopes Software: Astropy(AstropyCollaborationetal.2013); (MAST). STScI is operated by the Association of Uni- pyWWT (Robitaille et al. 2017)

APPENDIX A. CURATED DATA

Table 4. All Sky Surveys

Name (Bandpass) Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel)

Radio

Table 4 continued 14 Rosenfield et al. Table 4 (continued)

Name (Bandpass) Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel)

VLSS: VLA Low-frequency Sky Survey 79.2 Cohen et al.(2007) a VLA FIRST: Faint Images of the Radio Sky 9.90 Becker et al.(1995) at Twenty-centimeters NVSS: NRAO VLA Sky Survey 19.8 Condon et al.(1998) a SUMSS: Sydney University Molonglo Sky 19.8 Bock et al.(1999) (SUMSS, Min=0.001, Max=10). a Survey Westerbork Northern Sky Survey 9.90 WENSS Team. The WENSS project is a collaboration between the Netherlands Foundation for Research in Astronomy (NFRA/ASTRON) and the Leiden Observatory. (WENSS, Min=0.001, Max=10), who got it from the WENSS FTP site 1999-03-18. The original data was found using the Westerbork Synthesis Radio Telescope, which is operated by NFRA/ASTRON with support from the Netherlands Founda- tion for Scientiﬁc Research (NWO).a Bonn 1420 MHz Survey 39.6 Max Planck Institute for Radio Astronomy, generated by Pa- tricia Reich and Wolfgang Reich using with the Bonn Stockert 25m telescope. (1420MHz, Min=3, Max=35).a HI All-Sky Continuum Survey 158 Max Planck Institute for Radio Astronomy, generated by Glyn Haslam using data taken at Jodrell Bank, Eﬀelsberg and Parkes telescopes. (408MHz, Min=10, Max=200).a

Microwave

Planck 39.6 Planck is a European Space Agency mission, with signiﬁcant CMB participation from NASA. NASAs Planck Project Oﬃce is based at JPL. JPL contributed mission-enabling technology for Dust & Gas both of Plancks science instruments. European, Canadian and Thermal Dust U.S. Planck scientists work together to analyze the Planck data. Spinning Dust Molecular Gas (CO) Ionized Gas (free-free) Synchrotron (non-thermal) Lensing (Mass)

Table 4 continued WorldWide Telescope 15 Table 4 (continued)

Name (Bandpass) Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) WMAP 158 NASA/WMAP Science Team ILC 5-Year Cosmic Microwave Back- ground K Band, Linear, Non-Linear (23 GHz), and Polarization Map (50 uK) Ka Band, Linear, Non-Linear (33 GHz), and Polarization Map (35 uK) Q Band, Linear, Non-Linear (41 GHz), and Polarization Map (35 uK) V Band, Linear, Non-Linear (61 GHz), and Polarization Map (35 uK) W Band, Linear, Non-Linear (94 GHz), and Polarization Map (35 uK) WMAP QVW (Power and Linear)

Infrared

SFD Dust Map, 100 Micron 39.6 Data provided by two NASA satellites, the Infrared As- tronomy Satellite (IRAS) and the Cosmic Background Ex- plorer (COBE). Processing by David J. Schlegel, Douglas P. Finkbeiner and Marc Davis, Princeton University and Univer- sity of California, Berkeley. (SFD100m, Min=5, Max=5).a WISE All Sky 19.8 NASA/JPL-Caltech/UCLA 2MASS Two Micron All Sky Survey Syntheic catalog image by Jina Suh from 2MASS catalog Copy- Imagery right Microsoft 2007. Synthetic catalog 9.90

39.6

IRIS: Improved Reprocessing of IRAS Sur- 39.6 Red is IRIS100, Green IRIS60, Blue IRIS12 (IRIS12: Min=0.5, vey, 12, 25, 60, and 100 µm Max=1; IRIS25: Min=1, Max=2; IRIS60: Min=1, Max=2; IRIS100: Min=4, Max=4)ab COBE Diﬀuse Infrared Background Experi- 39.6 NASA/COBE (COBE, Min=10, Max=5).a ment (DIRBE) COBE DIRBE Annual Average Map 39.6 NASA/COBE (COBEAAM, Min=10, Max=5).a COBE DIRBE Zodi-Subtracted Mission 39.6 NASA/COBE (COBEZSMA, Min=10, Max=5).a Average

Optical

Table 4 continued 16 Rosenfield et al. Table 4 (continued)

Name (Bandpass) Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Hydrogen Alpha Full Sky Map 39.6 Image Courtesy Douglas Finkbeiner. The full-sky H-alpha map (6’ FWHM resolution) is a composite of the Virginia Tech Spec- tral line Survey (VTSS) in the north and the Southern H-Alpha Sky Survey Atlas (SHASSA) in the south. The Wisconsin H- Alpha Mapper (WHAM) survey provides a stable zero-point over 3/4 of the sky on a one degree scale. SDSS: Sloan Digital Sky Survey 0.30 Copyright SDSS Digitized Sky Survey 0.62 Copyright DSS Consortium Tycho (Synthetic) 79.2 NASA/Goddard Space Flight Center Scientiﬁc Visualization Studio USNOB: US Naval Observatory B 1.0 39.6 Monet(1998). Rendering by Jina Suh (Microsoft). (Synthetic)

Ultraviolet

GALEX 2 Combined 2.48 GR2/3 releaseae GALEX 2 Near-UV 2.48 GR2/3 releaseae GALEX 2 Far-UV 2.48 GR2/3 releaseae GALEX 4 Near-UV 2.48 Near UV band (1770-2730A)˚ of the GR4 releaseae GALEX 4 Far-UV 2.48 Far UV band (1350-1780A)˚ of the GR4 releaseae Extreme Ultraviolet Explorer (EUVE) 79.2 Red is 555 A(euve555),˚ Green 405 A,˚ Red 83 A.˚ ade EUVE: 83 A˚ 79.2 (euve83, Min=1, Max=150).ade EUVE: 171 A˚ 79.2 (euve171, Min=4, Max=200).ade EUVE: 405 A˚ 79.2 (euve405, Min=0.7, Max=70).ade EUVE: 555 A˚ 79.2 (euve555, Min=5, Max=500).ade

X-ray

RASS: ROSAT All Sky Survey 19.8 This is a composite of three RASS3 surveys from the ROSAT Data Archive of the Max-Planck-Institut fur extraterrestrische Physik (MPE) at Garching, Germany. Red is soft band (RASS3sb), Green is broad band (RASS3bb), Blue is hard band (RASS3hb)a ROSAT Hard Band Count Map 19.8 (RASS3hb, Min=0.2, Max=50).ac ROSAT Soft Band Count Map 19.8 (RASS3sb, Min=0.2, Max=50).ac ROSAT Broad Band Count Map 19.8 (RASS3bb, Min=0.2, Max=50).ac ROSAT Soft Band Intensity 19.8 (RASSSB, Min=0.2, Max=10).ac ROSAT Hard Band Intensity 19.8 (RASSHB, Min=0.2, Max=10).ac ROSAT PSPC summed pointed observations, 9.90 Observational data from NASA Goddard Space Flight Center, 2 degree cutoﬀ, intensity (PSPC2int, Min=0.02, Max=0.01).a

Table 4 continued WorldWide Telescope 17 Table 4 (continued)

Name (Bandpass) Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Swift BAT All-Sky Survey: Significance 14- 79.2 NASA BAT Team. (BATSig, Min=1, Max=10).a 195 keV Swift BAT All-Sky Survey: Flux 14-195 keV 79.2 NASA BAT Team. (BATFlux, Min=1e-5, Max=1e-3).a GRANAT/SIGMA Significance 79.2 (GRANAT SIGMA sig, Min=0.5, Max=50).af GRANAT/SIGMA Flux 79.2 (GRANAT SIGMA flux, Min=1e-5, Max=0.01).af

Gamma-ray

Fermi 158 NASA and the FERMI-LAT Team. Fermi Six Months 158 Fermi Year Two 39.6 NASA and the FERMI-LAT Team. Fermi Year Three 79.2 Fermi LAT Year Eight 39.6 CGRO Compton Telescope: 3 channel data 79.2 CompTel Instrument Team. Maps generated by Andrew Strong, Max-Planck Institute for Extraterrestrial Physics, Garching, Germany. (comptel, Min=0.05, Max=0.1).a EGRET Soft 79.2 EGRET Instrument team, NASA Goddard Space Flight Cen- ter. (EGRETsoft, Min=0.1, Max=0.002).a EGRET Hard 79.2 EGRET Instrument team, NASA Goddard Space Flight Cen- ter. (EGREThard, Min=0.02, Max=0.005).a aTOAST-formatted mosaics were obtained using facilities of NASAs SkyView Virtual Telescope. b Original IRAS data: NASA/JPL IPAC, IRIS Reprocessing: Canadian Institute for Theoretical Astrophysics/Institut d’Astrophysique Spatiale. c ROSAT Data Archive of the Max-Planck-Institut fur extraterrestrische Physik (MPE) at Garching, Germany. dCenter for Extreme UV Astronomy, University of California at Berkeley. e Data archived at MAST/STScI fHigh Energy Astrophysics Department, Space Research Institute, Moscow, Russia; CEA, Centre d’Etudes de Saclay Orme des Merisiers, France; Centre d’Etude Spatiale des Rayonnements, Toulouse, France; Federation de Recherche Astroparticule et Cosmologie Universite de Paris, France. 18 Rosenfield et al.

Table 5. Panoramas

Name Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Apollo 12: Landing Site 19.8 Panorama taken by Apollo 12 Lunar Module Pilot Alan Bean in 1969 and assembled at the NASA Ames Research Center in 2007. Images courtesy of NASA and the Lunar and Planetary Institute. Apollo 17: Shorty Crater 19.8 Panorama taken by Apollo 17 Commander Gene Cernan in 1972 and assembled at the NASA Ames Research Center in 2007. Images courtesy of NASA and the Lunar and Planetary Institute. Pathfinder: Improved MPF 360-degree Pres- 79.1 NASA/JPL. This is the 1993 “geometrically improved, color en- idential Panorama hanced” version of the 360-degree “Gallery Pan” at Ares Vallis, the first contiguous, uniform panorama taken by the Imager for Mars (IMP) over the course of Sols 8, 9, and 10. Pathfinder: “Many Rovers” 79.1 Panoramas made available by Dr. Carol Stoker, NASA Pathfinder: Monster (stereo) 79.1 AMES. NASA/JPL/USGS. Pathfinder: Landing site from Sagan Memo- 39.6 NASA/JPL/USGS rial Station (stereo) Spirit: McMurdo 19.8 This 360-degree view, called the “McMurdo” panorama, comes from the panoramic camera (Pancam) on NASA’s Mars Explo- ration Rover Spirit. From April through October 2006, Spirit has stayed on a small hill known as “Low Ridge.” There, the rover’s solar panels are tilted toward the sun to maintain enough solar power for Spirit to keep making scientific observations throughout the winter on southern Mars. This view of the surroundings from Spirit’s “Winter Haven” is presented in approximately true color. Image mosaicking: Kris Kapraro, Bob Deen, and the JPL/MIPL team.ac Spirit: McMurdo (false color) 19.8 The Pancam began shooting component images of this panorama during Spirit’s sol 814 (April 18, 2006) and completed the part shown here on sol 932 (Aug. 17, 2006). The panorama was acquired using all 13 of the Pancam’s color filters. Image mosaicking: Kris Kapraro, Bob Deen, and the JPL/MIPL team.ac Spirit: McMurdo (color stereo) 19.8 This 360-degree view comes from the panoramic camera (Pan- cam) on NASA’s Mars Exploration Rover Spirit. From April through October 2006, Spirit has stayed on a small hill known as “Low Ridge.” Spirit: McMurdo (left eye) 19.8 The Pancam on the Spirit Rover began shooting component Spirit: McMurdo (right eye) 19.8 images of this panorama during Spirit’s sol 814 (April 18, 2006) and completed the part shown here on sol 932 (Aug. 17, 2006). The panorama was acquired using all 13 of the Pancam’s color filters, using lossless compression for the red and blue stereo filters, and only modest levels of compression on the remaining filters. The overall panorama consists of 1,449 Pancam images.a

Spirit: West Valley 19.8 NASA’S Mars Exploration Rover Spirit captured this westward view from atop a low plateau where Sprit spent the closing months of 2007. NASA/JPL-Caltech/Cornell Table 5 continuedUniversity. WorldWide Telescope 19 Table 5 (continued)

Name Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Spirit: West Valley (false color) 19.8 Spirit: West Valley (stereo) 19.8 tablebreak Spirit: Descent from Husband Hill 19.8 Spirit acquired the 405 individual images that make up this 360-degree view of the surrounding terrain using five different Spirit: Descent from Husband Hill (false 19.8 filters on the panoramic camera. The rover took the images color) on Martian days, or sols, 672 to 677 (Nov. 23 to 28, 2005). abc Spirit: Descent from Husband Hill (stereo) 19.8 Spirit: Gibson Panorama at Home Plate 19.8 NASA’s Mars Exploration Rover Spirit acquired this (false color) high-resolution view of intricately layered exposures of rock while parked on the northwest edge of the bright, Spirit: Gibson Panorama at Home Plate 39.6 semi-circular feature known as “Home Plate.” The rover was perched at a 27-degree upward tilt while creating the panorama, resulting in the “U” shape of the mosaic. d Spirit: Home Plate South 19.8 This is the Spirit Pancam “Home Plate South” panorama, acquired on sols 1325 - 1332 (September 25 - October 2, 2007). This is an approximate true color rendering using Pancam’s 753 nm, 535 nm, and 432 nm filters.a Spirit: Home Plate South (false color) 19.8 This is a false-color version of the Spirit Pancam “Home Plate South” panorama, acquired on sols 1325 - 1332 (September 25 Spirit: Home Plate South (stereo) 19.8 - October 2, 2007).a Spirit: Husband Hill Summit 39.6 The panoramic camera on NASA’s Mars Exploration Rover Spirit took the hundreds of images combined into this 360- degree view, the “Husband Hill Summit” panorama. The images were acquired on Spirit’s sols 583 to 586 (Aug. 24 to 27, 2005), shortly after the rover reached the crest of “Husband Hill” inside Mars’ Gusev Crater. Spirit: Santa Anita 19.8 This color mosaic was taken on May 21, 25 and 26, 2004, by the panoramic camera on NASA’s Mars Exploration Rover Spirit was acquired from a position roughly three-fourths the way between “Bonneville Crater” and the base of the “Columbia Hills”.a Spirit: Independence 19.8 The 108 images used to make this panorama were obtained on sols 536-543 (July 6 to 13, 2005) from a position in the Columbia Hills near the summit of Husband Hill. Spirit: Everest 19.8 It took Spirit three days, sols 620 to 622 (Oct. 1 to Oct. 3, 2005), to acquire all the 81 images combined into this mosaic, looking outward in every direction from the true summit of Husband Hill. Sky fixed by Jim St George.abc Spirit: Everest (stereo) 19.8 Composed of 81 images acquired on sols 620 to 622 (October 1 to 3, 2005) from a position in the Columbia Hills at the true summit of Husband Hill.ab Spirit Mission Success 39.6 This panorama was acquired in eight parts called octants, over the course of three different sols. Because the octants were taken at different times of day and across different days when the dust abundance was changing, there were large brightness and color seams between the octants in the assembled mosaic.a Spirit: Thanksgiving 19.8 This is the Spirit Pancam “Thanksgiving” panorama, acquired on sols 318 to 325 (Nov. 24 to Dec. 2, 2004) from a position along the flank of Husband Hill, which is the peak Table 5 continuedjust left of the center of this mosaic, just east of the West Spur of the Columbia Hills.abc 20 Rosenfield et al. Table 5 (continued)

Name Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Spirit: Thanksgiving (stereo) 79.1

tablebreak Spirit: Cahokia 19.8 The approximate true-color image, nicknamed the “Cahokia Spirit: Cahokia (stereo, unadjusted) 19.8 panorama” after the Native American archaeological site near Spirit: Cahokia (stereo, tilt-adjusted) 158 St. Louis, was acquired between sols 213 to 223 (Aug. 9 to 19, 2004). The panorama consists of 470 images acquired through six panoramic camera ﬁlters (750 to 480 nanometers). Stereo images: JPL/MIPL (Bob Deen).a

Spirit: Larry’s Lookout 19.8 This is the Spirit Pancam “Lookout” panorama, acquired on Spirit: Larry’s Lookout (stereo) 19.8 sols 410 to 413 (Feb. 27 to Mar. 2, 2005) from a position known informally as “Larry’s Lookout” along the drive up Husband Hill.abc

Spirit: Whale of a Panorama 19.8 Spirit produced this 220 degree image mosaic two-thirds of the way to the summit of Husband Hill from images collected from sol 497 to 500 (May 27 through May 30, 2005).a Spirit: From The Summit 39.6 This approximate true-color 240 degree panorama was taken by NASA’s Spirit rover from the top of “Husband Hill” in the “Columbia Hills” of Gusev Crater. The mosaic is made up of images taken by the rover’s panoramic camera over a period of three days (sols 583 to 585, or August 24 to 26, 2005). Spirit: Paige Panorama 39.6 This 230-degree panorama was composed from 72 images from d Spirit: Paige Panorama (false color) 39.6 the Pancam on Spirit on Feb 19 2006. Spirit: Legacy 19.8 The 78 images used to make this were acquired on sols 59 to 61 (March 3 to 5, 2004) from a position about halfway between the landing site and the rim of Bonneville crater, within the transition from the relatively smooth plains to the more rocky and rugged ejecta blanket of Bonneville.abc Spirit: Bonneville 19.8 The rim and interior of a crater nicknamed “Bonneville” dom- inate this 180-degree, false-color mosaic. Spirit recorded this view on the rover’s 68th sol, March 12, 2004, one sol after reaching this location. The rover remained here in part to get this very high-resolution, color mosaic, from which scientists can gain insight about the depth of the surface material at Bon- neville and make future observation plans. On sol 71, Spirit was instructed to drive approximately 15 meters (49 feet) along the crater rim to a new vantage point.a Opportunity: Erebus 19.8 This panorama was made from 635 images acquired on sols 652 to 663 (Nov. 23 to Dec. 5, 2005), as NASA’s Mars Opportunity: Erebus (Stereo) 19.8 Exploration Rover Opportunity was exploring sand dunes and outcrop rocks in Meridiani Planum. Image mosaicking: JPL/MIPL Team (Bob Deen, Oleg Pariser, Jeﬀrey Hall)ac.

Table 5 continued WorldWide Telescope 21 Table 5 (continued)

Name Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) tablebreak Opportunity: Lyell 39.6 This view combines many images taken by Opportunity’s panoramic camera (Pancam) from the 1,332nd through Opportunity: Lyell (Stereo) 19.8 1,379th Martian days, or sols, of the mission (Oct. 23 to Dec. 11, 2007). The main body of the crater appears in the upper right of this stereo panorama, with the far side of the crater lying about 800 meters (half a mile) away. Bracketing that part of the view are two promontories on the crater’s rim at either side of Duck Bay. They are “Cape Verde,” about 6 meters (20 feet) tall, on the left, and “Cabo Frio,” about 15 meters (50 feet) tall, on the right. The rest of the image, other than sky and portions of the rover, is ground within Duck Bay. Image mosaicking: Jim Bell and Jonathan Joseph.ac Opportunity: Rub al Khali 19.8 Taken on the plains of Meridiani during the period from the rover’s 456th to 464th sols on Mars (May 6 to May 14, 2005). Opportunity was about 2 kilometers (1.2 miles) south of “En- durance Crater” at a place known informally as “Purgatory Dune”.a tablebreak Opportunity: Endurance South 19.8 This 360-degree panorama shows “Endurance Crater” and the surrounding plains of Meridiani Planum. This is the second Opportunity: Endurance South (false color) 19.8 large panoramic camera mosaic of Endurance, and was obtained from a high point near the crater’s south rim.a Opportunity Mission Succcess 19.8 This expansive view of the martian real estate surrounding the Mars Exploration Rover Opportunity is the first 360 degree, high-resolution color image taken by the rover’s panoramic camera. The airbag marks, or footprints, seen in the soil trace the route by which Opportunity rolled to its final resting spot inside a small crater at Meridiani Planum, Mars. The exposed rock outcropping is a future target for further examination. This image mosaic consists of 225 individual frames. Opportunity: Burns Cliff 39.6 This view of “Burns Cliff” after driving right to the base of this southeastern portion of the inner wall of “Endurance Crater.” The view combines 46 images taken by Opportunity’s panoramic camera between the rover’s 287th and 294th martian days (Nov. 13 to 20, 2004).a Opportunity: Payson Panorama 79.1 The panoramic camera aboard NASA’s Mars Exploration Rover Opportunity acquired this panorama of the Payson outcrop on the western edge of Erebus Crater during Opportunity’s sol 744 (Feb. 26, 2006). USGSa. Opportunity: Beagle Crater 19.8 Opportunity took the mosaic of images that make up this 360-degree view of the rover’s surroundings with the Opportunity: Beagle Crater (false color) 19.8 panoramic camera on the rover’s 901st through 904th sols, or Martian days (Aug. 6 through Aug. 9, 2006), of exploration. This is an approximate true-color image combining exposures taken through the panoramic camera’s 753-nanometer, 535-namometer, and 432-nanometer filters. Image mosaicking: Cornell Pancam team. NASA/JPL/Cornell/UNMc

Table 5 continued 22 Rosenfield et al. Table 5 (continued)

Name Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Opportunity: Lion King 19.8 This shows “Eagle Crater” and the surrounding plains of Merid- iani Planum. It was stitched from 558 images obtained on sols 58 and 60 by the Mars Exploration Rover Opportunity’s panoramic camera.a Phoenix: Landing Site 39.6 This view combines more than 400 images taken during the ﬁrst several weeks after NASA’s Phoenix Mars Lander arrived on an arctic plain at 68.22 degrees north latitude, 234.25 degrees east longitude on Mars. NASA/JPL-Caltech/University Arizona/Texas A & M University. aNASA/JPL/Cornell. b Image mosaicking: JPL/MIPL (Bob Deen) c Calibration and color rendering: Cornell Calibration Crew and the Pancam team (Jim Bell) dNASA/JPL-Caltech/USGS/Cornell.

Table 6. Planet Maps

Name Approximate Notes, Credits

TOAST Resolution (arcsec/pixel) Bing Maps Aerial 0.002 Bing Maps Hybrid 0.002 Bing Maps Copyright Microsoft Corp. and Suppliers Bing Maps Streets 0.002 Earth at Night 2012 4.94 NASA Earth Observatory/NOAA NGDC Earth at Night 19.8 Data courtesy Marc Imhoﬀ of NASA GSFC and Christopher Elvidge of NOAA NGDC. Image by Craig Mayhew and Robert Simmon, NASA GSFC. Blue Marble January 2004 19.8 Blue Marble Next Generation, January 2004. Visible Earth Project, NASA. Blue Marble July 2004 19.8 Blue Marble Next Generation, July 2004. Visible Earth Project, NASA. Open Street Maps 0.010 Map data Copyright OpenStreetMap contributors, CC BY-SA Open Street Maps Bike Map 0.010 Map data Copyright OpenStreetMap contributors, CC BY-SA Moon 2.47 Image Courtesy NASA/JPL Mercury 9.89 NASA/Johns Hopkins University Applied Physics Labora- tory/Carnegie Institution of Washington Venus 79.1 Image Courtesy NASA/JPL/Space Science Institute Mars Visible Imagery 9.89 NASA/USGS/Malin Space Science Systems/JPL

Table 6 continued WorldWide Telescope 23 Table 6 (continued)

Name Approximate Notes, Credits

TOAST Resolution

(arcsec/pixel) Jupiter 158 Image Courtesy NASA/JPL/Space Science Institute Phobos (Mars) 316 Viking images of the moon Phobos.a Deimos (Mars) 316 Viking images of the moon Deimos.a Io (Jupiter) 79.1 Europa (Jupiter) 79.1 Images Courtesy NASA/JPL/Space Science Institute. Ganymede (Jupiter) 79.1 Callisto (Jupiter) 79.1 Mimas (Saturn) 316 Approximately half the surface of moon Mimasab Enceladus (Saturn) 316 Approximately half the surface of moon Enceladusab Tethys (Saturn) 316 Most of the surface of the moon Tethysab Dione (Saturn) 316 Most of the surface of the moon Dioneab Rhea (Saturn) 316 The moon Rheaab Iapetus (Saturn) 316 Approximately half of the surface of the moon Iapetusab Ariel (Uranus) 316 Most of the Southern Hemisphere of the moon Arielab Umbriel (Uranus) 316 Most of the Southern Hemisphere of the moon Umbrielab Titania (Uranus) 316 Most of the Southern Hemisphere of the moon Titaniaab Oberon (Uranus) 316 Most of the Southern Hemisphere of the moon Oberonab Miranda (Uranus) 316 Most of the Southern Hemisphere of the moon Mirandaab aCaltech/JPL/USGS. b From Voyager mission photographs.

REFERENCES

Acton, C., Bachman, N., Semenov, B., & Wright, E. 2017, Condon, J. J., Cotton, W. D., Greisen, E. W., et al. 1998, Planetary and Space Science AJ, 115, 1693 Acton, C. H. 1996, Planet. Space Sci., 44, 65 Cui, C., & Li, S. 2018, ArXiv e-prints, arXiv:1801.05098 Agarwal, D., Cheah, Y.-W., Fay, D., et al. 2011, The Currie, T., Lisse, C. M., Kuchner, M., et al. 2015, ApJL, International Journal of High Performance Computing 807, L7 Applications, 25, 304. Geller, M. J., & Huchra, J. P. 1989, Science, 246, 897 https://doi.org/10.1177/1094342011414746 Gray, J., & Szalay, A. S. 2004, eprint arXiv:cs/0403018, Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., cs/0403018 et al. 2013, A&A, 558, A33 Jurić,M., Kantor, J., Lim, K., et al. 2015, ArXiv e-prints, Batygin, K., & Brown, M. E. 2016, AJ, 151, 22 arXiv:1512.07914 Becker, R. H., White, R. L., & Helfand, D. J. 1995, ApJ, Kluyver, T., Ragan-Kelley, B., Pérez,F., et al. 2016, in 450, 559 ELPUB, 87–90 Berriman, G. B., & Good, J. C. 2017, PASP, 129, 058006 Kunszt, P. Z., Szalay, A. S., Csabai, I., & Thakar, A. R. Bock, D. C.-J., Large, M. I., & Sadler, E. M. 1999, AJ, 117, 2000, in Astronomical Society of the Pacific Conference 1578 Series, Vol. 216, Astronomical Data Analysis Software Cohen, A. S., Lane, W. M., Cotton, W. D., et al. 2007, AJ, and Systems IX, ed. N. Manset, C. Veillet, & 134, 1245 D. Crabtree, 141 24 Rosenfield et al.

Ladd, E. F., Gingrich, E. C., Nottis, K. E. K., Sanders, N. E., Faesi, C., & Goodman, A. A. 2014, Journal Udomprasert, P., & Goodman, A. A. 2015, in of Science Education and Technology, 23, 431 Astronomical Society of the Pacific Conference Series, Scoville, N., Aussel, H., Brusa, M., et al. 2007, ApJS, 172, 1 Vol. 500, Celebrating Science: Putting Education Best Practices to Work, ed. G. Shultz, S. Buxner, L. Shore, & Udomprasert, P., Goodman, A., Sunbury, S., et al. 2014, in J. Barnes, 191 Astronomical Society of the Pacific Conference Series, McKinney, W. 2010, in Proceedings of the 9th Python in Vol. 483, Ensuring Stem Literacy: A National Conference Science Conference, ed. S. van der Walt & J. Millman, 51 on STEM Education and Public Outreach, ed. J. G. – 56 Manning, M. K. Hemenway, J. B. Jensen, & M. G. Monet, D. 1998, USNO-A2.0 (United States Naval Gibbs, 297 Observatory.) Nakata, M., Hamacher, D. W., Warren, J., et al. 2014, Udomprasert, P. S., Goodman, A. A., Plummer, J., et al. Australian Academic & Research Libraries, Vol. 45(2), 2016, in American Astronomical Society Meeting pp. 101-110., 45, 101 Abstracts, Vol. 227, American Astronomical Society Pérez,F., & Granger, B. E. 2007, Computing in Science Meeting Abstracts, 229.01 and Engineering, 9, 21. http://ipython.org Udomprasert, P. S., Goodman, A. A., & Wong, C. 2012, in Perryman, M. A. C., Lindegren, L., Kovalevsky, J., et al. 1997, A&A, 323, L49 Astronomical Society of the Pacific Conference Series, Qiao, C., Cui, C., Zheng, X., & Xu, Y. 2010, in 2010 2nd Vol. 457, Connecting People to Science: A National International Conference on Education Technology and Conference on Science Education and Public Outreach, Computer, Vol. 3, V3–519–V3–523 ed. J. B. Jensen, J. G. Manning, M. G. Gibbs, & Roberts, M. J. 2014, in Astronomical Society of the Pacific D. Daou, 149 Conference Series, Vol. 483, Ensuring Stem Literacy: A Wang, Q., Qiao, C., & Zheng, X. 2015, International National Conference on STEM Education and Public Outreach, ed. J. G. Manning, M. K. Hemenway, J. B. Journal of Learning and Teaching, 1, 134 Jensen, & M. G. Gibbs, 223 Wong, C. 2008, in Astronomical Society of the Pacific Robitaille, T., Otor, O. J., ZuHone, J., & Rosenfield, P. Conference Series, Vol. 389, EPO and a Changing World: 2017, PyWWT: WorldWide Telescope from Python, , , Creating Linkages and Expanding Partnerships, ed. doi:10.5281/zenodo.1155830. C. Garmany & M. G. Gibbs, Moody J. W., 227 https://doi.org/10.5281/zenodo.1155830 Rosenfield, P., Gaily, J., Fraser, O., & Wisniewski, J. 2014, Yusef-Zadeh, F., Roberts, D. A., Wardle, M., et al. 2015, Communicating Astronomy with the Public Journal, 15, ApJL, 801, L26 35